Unlike whole genome-wide association studies (GWAS) which can only discover variations in DNA that are frequent in the populations (greater than 1%), both exome sequencing and whole-genome sequencing may have the advantage of testing comprehensively the role of coding variation, both common and rare. It is expected that every gene may harbor functionally relevant variants. We are developing a new method to address the issues of population stratification. Using the Latino/Hispanic studies as an example, while carrying out common or rare variants analysis, it is important to consider the ethnicity impacts. These methods/pipelines are under development. We will soon start to apply our methods on real data including obsessive compulsive disorder, bipolar and schizophrenia.